Beyond the Hype: How Scientific Companies Actually Feel About AI
Everyone's using AI. Almost no one trusts it.
We recently surveyed scientific and B2B technology companies to understand how they're actually using AI today.
Not the hype. Not the headlines. The reality.
The results confirmed something we've been sensing in conversations with clients and prospects: everyone's experimenting, but almost no one fully trusts it yet.
The AI Trust Gap
61% of respondents shared that they are actively using or piloting AI in their business. That's not surprising: the tools are accessible, the pressure to adopt is real, and most leaders know they need to be doing something.
But here's where it gets interesting.
Only 1.6% said they have high confidence in AI and fully rely on it. One person out of 64.
Meanwhile, 64% described themselves as "cautiously optimistic"—they're using AI, but keeping humans in the loop because the error rate is still too high to let go of the wheel.
The access problem is solved. The trust problem isn't.
What's Actually Holding Companies Back?
We asked what's preventing organizations from adopting AI more fully. The top barriers weren't what you might expect.
- Lack of internal expertise: 20%
- Trust/accuracy concerns: 19%
- Data privacy or security concerns: 19%
- Leadership hasn't prioritized it: 17%
Cost and budget? Only 6% cited that as the main barrier.
The tools aren't expensive anymore. What's missing is the know-how, the confidence, and in many cases, the leadership commitment to figure it out.
Marketing Sees the Wave Coming
When we asked which business function AI will disrupt most, and soonest, the answer was overwhelming:
- Marketing & Content: 67%
- Software Development: 48%
- Customer Support: 38%
- Sales & Business Development: 38%
The people in this survey (largely sales and marketing professionals) see their own roles as ground zero for disruption. That's either self-awareness or anxiety. Probably both.
But when we looked at how they're actually using AI today, the picture is more modest:
- Writing and editing emails/documents: 70%
- Research and summarization: 53%
- Content creation and copywriting: 44%
- Meeting transcription: 39%
- Brainstorming and ideation: 34%
These are acceleration tasks. AI is making existing work faster, not transforming how the work gets done.
Most companies are using AI as a faster pen, not rethinking the page.
The "Promising, But..." Majority
Perhaps the most telling finding: when asked about AI's practical value today, only 33% said it's already delivering real ROI.
45% said it's "promising, but not quite there yet."
That's the middle most companies are stuck in. They see the potential and they're experimenting, but they haven't cracked the code on turning experimentation into outcomes.
What This Means
AI adoption isn't a question of access anymore. The tools are here. The costs have dropped. The barrier now is implementation—knowing where AI actually creates value, having the expertise to deploy it properly, and building enough trust to change how work gets done.
The companies that figure this out won't just move faster. They'll operate on a different level entirely.
